A generalized Chua cell for realizing any continuous n-segment piecewise-linear function

In this letter it is proven that any piecewise-linear (PWL) function can be realized as the sum of a definite number of basic PWL functions, like that implemented in the classical Cellular Neural Network (CNN) structure by L. Chua. This allows us to consider the basic Chua's PWL function as a fundamental unit to build up a generic PWL function and, thus, to extend the class of nonlinear dynamics that can be realized by CNNs to a more wide class including PWL approximation of any nonlinearity. Moreover, an example of the cell realization of such a CNN is reported.